3 from __future__ import unicode_literals
8 from .common import InfoExtractor
9 from .youtube import YoutubeIE
13 compat_urllib_request,
15 compat_xml_parse_error,
25 from .brightcove import BrightcoveIE
26 from .ooyala import OoyalaIE
27 from .rutv import RUTVIE
28 from .smotri import SmotriIE
31 class GenericIE(InfoExtractor):
32 IE_DESC = 'Generic downloader that works on some sites'
37 'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
38 'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
40 'id': '13601338388002',
42 'uploader': 'www.hodiho.fr',
43 'title': 'R\u00e9gis plante sa Jeep',
46 # bandcamp page with custom domain
48 'add_ie': ['Bandcamp'],
49 'url': 'http://bronyrock.com/track/the-pony-mash',
53 'title': 'The Pony Mash',
54 'uploader': 'M_Pallante',
56 'skip': 'There is a limit of 200 free downloads / month for the test song',
58 # embedded brightcove video
59 # it also tests brightcove videos that need to set the 'Referer' in the
62 'add_ie': ['Brightcove'],
63 'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
65 'id': '2765128793001',
67 'title': 'Le cours de bourse : l’analyse technique',
68 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
69 'uploader': 'BFM BUSINESS',
72 'skip_download': True,
76 # https://github.com/rg3/youtube-dl/issues/2253
77 'url': 'http://bcove.me/i6nfkrc3',
78 'md5': '0ba9446db037002366bab3b3eb30c88c',
80 'id': '3101154703001',
82 'title': 'Still no power',
83 'uploader': 'thestar.com',
84 'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
86 'add_ie': ['Brightcove'],
89 'url': 'http://www.championat.com/video/football/v/87/87499.html',
90 'md5': 'fb973ecf6e4a78a67453647444222983',
92 'id': '3414141473001',
94 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
95 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
96 'uploader': 'Championat',
99 # Direct link to a video
101 'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
102 'md5': '67d406c2bcb6af27fa886f31aa934bbe',
107 'upload_date': '20100513',
112 'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
113 'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
115 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
117 'title': '2cc213299525360.mov', # that's what we get
122 'url': 'http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCUQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcmQHVoWB5FY&ei=F-sNU-LLCaXk4QT52ICQBQ&usg=AFQjCNEw4hL29zgOohLXvpJ-Bdh2bils1Q&bvm=bv.61965928,d.bGE',
126 'upload_date': '20130224',
127 'uploader_id': 'TheVerge',
128 'description': 'Chris Ziegler takes a look at the Alcatel OneTouch Fire and the ZTE Open; two of the first Firefox OS handsets to be officially announced.',
129 'uploader': 'The Verge',
130 'title': 'First Firefox OS phones side-by-side',
133 'skip_download': False,
138 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
142 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
143 'upload_date': '20140225',
144 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
145 'uploader': 'Tested',
146 'uploader_id': 'testedcom',
148 # No need to test YoutubeIE here
150 'skip_download': True,
155 'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
156 'md5': '7cf780be104d40fea7bae52eed4a470e',
160 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
161 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
166 'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
170 'title': 'Охотское море стало целиком российским',
171 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
175 'skip_download': True,
180 'url': 'http://en.support.wordpress.com/videos/ted-talks/',
181 'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
185 'title': 'My web playroom',
186 'uploader': 'Ze Frank',
187 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
190 # Embeded Ustream video
192 'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
193 'md5': '27b99cdb639c9b12a79bca876a073417',
197 'uploader': 'AU SPA: The NSA and Privacy',
198 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
201 # nowvideo embed hidden behind percent encoding
203 'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
204 'md5': '2baf4ddd70f697d94b1c18cf796d5107',
206 'id': '06e53103ca9aa',
208 'title': 'Macross Episode 001 Watch Macross Episode 001 onl',
209 'description': 'No description',
214 'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
215 'md5': '7653032cbb25bf6c80d80f217055fa43',
217 'id': '048195-004_PLUS7-F',
220 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
221 'upload_date': '20140320',
224 'skip_download': 'Requires rtmpdump'
229 'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
230 'md5': 'ec40048448e9284c9a1de77bb188108b',
232 'id': 'v27008541fad',
234 'title': 'Крым и Севастополь вошли в состав России',
235 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
237 'upload_date': '20140318',
238 'uploader': 'rbctv_2012_4',
239 'uploader_id': 'rbctv_2012_4',
244 'url': 'http://www.wired.com/2014/04/honda-asimo/',
245 'md5': 'ba0dfe966fa007657bd1443ee672db0f',
247 'id': '53501be369702d3275860000',
249 'title': 'Honda’s New Asimo Robot Is More Human Than Ever',
254 'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
255 'md5': '441aeeb82eb72c422c7f14ec533999cd',
257 'id': 'k2mm4bCdJ6CQ2i7c8o2',
259 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
262 'add_ie': ['Dailymotion'],
264 # YouTube embed via <data-embed-url="">
266 'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
267 'md5': 'c267b1ab6d736057d64babaa37e07a66',
271 'title': 'Asphalt 8: Airborne - Chinese Great Wall - Android Game Trailer',
272 'uploader': 'gameloftandroid',
273 'uploader_id': 'gameloftandroid',
274 'upload_date': '20140321',
275 'description': 'md5:9c6dca5dd75b7131ce482ccf080749d6'
280 def report_download_webpage(self, video_id):
281 """Report webpage download."""
282 if not self._downloader.params.get('test', False):
283 self._downloader.report_warning('Falling back on generic information extractor.')
284 super(GenericIE, self).report_download_webpage(video_id)
286 def report_following_redirect(self, new_url):
287 """Report information extraction."""
288 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
290 def _send_head(self, url):
291 """Check if it is a redirect, like url shorteners, in case return the new url."""
293 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
295 Subclass the HTTPRedirectHandler to make it use our
296 HEADRequest also on the redirected URL
298 def redirect_request(self, req, fp, code, msg, headers, newurl):
299 if code in (301, 302, 303, 307):
300 newurl = newurl.replace(' ', '%20')
301 newheaders = dict((k,v) for k,v in req.headers.items()
302 if k.lower() not in ("content-length", "content-type"))
304 # This function was deprecated in python 3.3 and removed in 3.4
305 origin_req_host = req.get_origin_req_host()
306 except AttributeError:
307 origin_req_host = req.origin_req_host
308 return HEADRequest(newurl,
310 origin_req_host=origin_req_host,
313 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
315 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
317 Fallback to GET if HEAD is not allowed (405 HTTP error)
319 def http_error_405(self, req, fp, code, msg, headers):
323 newheaders = dict((k,v) for k,v in req.headers.items()
324 if k.lower() not in ("content-length", "content-type"))
325 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
327 origin_req_host=req.get_origin_req_host(),
331 opener = compat_urllib_request.OpenerDirector()
332 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
333 HTTPMethodFallback, HEADRedirectHandler,
334 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
335 opener.add_handler(handler())
337 response = opener.open(HEADRequest(url))
339 raise ExtractorError('Invalid URL protocol')
342 def _extract_rss(self, url, video_id, doc):
343 playlist_title = doc.find('./channel/title').text
344 playlist_desc_el = doc.find('./channel/description')
345 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
349 'url': e.find('link').text,
350 'title': e.find('title').text,
351 } for e in doc.findall('./channel/item')]
356 'title': playlist_title,
357 'description': playlist_desc,
361 def _real_extract(self, url):
362 if url.startswith('//'):
365 'url': self.http_scheme() + url,
368 parsed_url = compat_urlparse.urlparse(url)
369 if not parsed_url.scheme:
370 default_search = self._downloader.params.get('default_search')
371 if default_search is None:
372 default_search = 'auto_warning'
374 if default_search in ('auto', 'auto_warning'):
376 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
377 return self.url_result('http://' + url)
379 if default_search == 'auto_warning':
380 self._downloader.report_warning(
381 'Falling back to youtube search for %s . Set --default-search to "auto" to suppress this warning.' % url)
382 return self.url_result('ytsearch:' + url)
384 assert ':' in default_search
385 return self.url_result(default_search + url)
386 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
388 self.to_screen('%s: Requesting header' % video_id)
391 response = self._send_head(url)
394 new_url = response.geturl()
396 self.report_following_redirect(new_url)
397 return self.url_result(new_url)
399 # Check for direct link to a video
400 content_type = response.headers.get('Content-Type', '')
401 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
403 upload_date = response.headers.get('Last-Modified')
405 upload_date = unified_strdate(upload_date)
408 'title': os.path.splitext(url_basename(url))[0],
410 'format_id': m.group('format_id'),
412 'vcodec': 'none' if m.group('type') == 'audio' else None
414 'upload_date': upload_date,
417 except compat_urllib_error.HTTPError:
418 # This may be a stupid server that doesn't like HEAD, our UA, or so
422 webpage = self._download_webpage(url, video_id)
424 # since this is the last-resort InfoExtractor, if
425 # this error is thrown, it'll be thrown here
426 raise ExtractorError('Failed to download URL: %s' % url)
428 self.report_extraction(video_id)
432 doc = parse_xml(webpage)
434 return self._extract_rss(url, video_id, doc)
435 except compat_xml_parse_error:
438 # Sometimes embedded video player is hidden behind percent encoding
439 # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
440 # Unescaping the whole page allows to handle those cases in a generic way
441 webpage = compat_urllib_parse.unquote(webpage)
443 # it's tempting to parse this further, but you would
444 # have to take into account all the variations like
445 # Video Title - Site Name
446 # Site Name | Video Title
447 # Video Title - Tagline | Site Name
448 # and so on and so forth; it's just not practical
449 video_title = self._html_search_regex(
450 r'(?s)<title>(.*?)</title>', webpage, 'video title',
453 # video uploader is domain name
454 video_uploader = self._search_regex(
455 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
457 # Look for BrightCove:
458 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
460 self.to_screen('Brightcove video detected.')
463 'url': smuggle_url(bc_url, {'Referer': url}),
464 'ie_key': 'Brightcove'
465 } for bc_url in bc_urls]
469 'title': video_title,
474 # Look for embedded (iframe) Vimeo player
476 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
478 player_url = unescapeHTML(mobj.group('url'))
479 surl = smuggle_url(player_url, {'Referer': url})
480 return self.url_result(surl, 'Vimeo')
482 # Look for embedded (swf embed) Vimeo player
484 r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
486 return self.url_result(mobj.group(1), 'Vimeo')
488 # Look for embedded YouTube player
489 matches = re.findall(r'''(?x)
490 (?:<iframe[^>]+?src=|data-video-url=|embedSWF\(\s*)
491 (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
495 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
496 for tuppl in matches]
497 return self.playlist_result(
498 urlrs, playlist_id=video_id, playlist_title=video_title)
500 # Look for embedded Dailymotion player
501 matches = re.findall(
502 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
504 urlrs = [self.url_result(unescapeHTML(tuppl[1]))
505 for tuppl in matches]
506 return self.playlist_result(
507 urlrs, playlist_id=video_id, playlist_title=video_title)
509 # Look for embedded Wistia player
511 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
514 '_type': 'url_transparent',
515 'url': unescapeHTML(match.group('url')),
517 'uploader': video_uploader,
518 'title': video_title,
522 # Look for embedded blip.tv player
523 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
525 return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
526 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
528 return self.url_result(mobj.group(1), 'BlipTV')
530 # Look for embedded condenast player
531 matches = re.findall(
532 r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
539 'ie_key': 'CondeNast',
541 } for ma in matches],
542 'title': video_title,
546 # Look for Bandcamp pages with custom domain
547 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
549 burl = unescapeHTML(mobj.group(1))
550 # Don't set the extractor because it can be a track url or an album
551 return self.url_result(burl)
553 # Look for embedded Vevo player
555 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
557 return self.url_result(mobj.group('url'))
559 # Look for Ooyala videos
560 mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
561 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
563 return OoyalaIE._build_url_result(mobj.group('ec'))
565 # Look for Aparat videos
566 mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
568 return self.url_result(mobj.group(1), 'Aparat')
570 # Look for MPORA videos
571 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
573 return self.url_result(mobj.group(1), 'Mpora')
575 # Look for embedded NovaMov-based player
577 r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
578 (?P<url>http://(?:(?:embed|www)\.)?
580 nowvideo\.(?:ch|sx|eu|at|ag|co)|
581 videoweed\.(?:es|com)|
582 movshare\.(?:net|sx|ag)|
583 divxstage\.(?:eu|net|ch|co|at|ag))
584 /embed\.php.+?)\1''', webpage)
586 return self.url_result(mobj.group('url'))
588 # Look for embedded Facebook player
590 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
592 return self.url_result(mobj.group('url'), 'Facebook')
594 # Look for embedded VK player
595 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
597 return self.url_result(mobj.group('url'), 'VK')
599 # Look for embedded Huffington Post player
601 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
603 return self.url_result(mobj.group('url'), 'HuffPost')
606 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
608 return self.url_result(mobj.group('url'))
609 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
611 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
613 # Look for funnyordie embed
614 matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
616 urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
618 return self.playlist_result(
619 urlrs, playlist_id=video_id, playlist_title=video_title)
621 # Look for embedded RUTV player
622 rutv_url = RUTVIE._extract_url(webpage)
624 return self.url_result(rutv_url, 'RUTV')
626 # Look for embedded TED player
628 r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
630 return self.url_result(mobj.group('url'), 'TED')
632 # Look for embedded Ustream videos
634 r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
636 return self.url_result(mobj.group('url'), 'Ustream')
638 # Look for embedded arte.tv player
640 r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
643 return self.url_result(mobj.group('url'), 'ArteTVEmbed')
645 # Look for embedded smotri.com player
646 smotri_url = SmotriIE._extract_url(webpage)
648 return self.url_result(smotri_url, 'Smotri')
650 # Look for embeded soundcloud player
652 r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
655 url = unescapeHTML(mobj.group('url'))
656 return self.url_result(url)
658 # Start with something easy: JW Player in SWFObject
659 found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
661 # Look for gorilla-vid style embedding
662 found = re.findall(r'''(?sx)
666 jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
668 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
670 # Broaden the search a little bit
671 found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
673 # Broaden the findall a little bit: JWPlayer JS loader
674 found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
676 # Try to find twitter cards info
677 found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
679 # We look for Open Graph info:
680 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
681 m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
682 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
683 if m_video_type is not None:
684 found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
687 found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
690 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
691 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
694 new_url = found.group(1)
695 self.report_following_redirect(new_url)
701 raise ExtractorError('Unsupported URL: %s' % url)
704 for video_url in found:
705 video_url = compat_urlparse.urljoin(url, video_url)
706 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
708 # Sometimes, jwplayer extraction will result in a YouTube URL
709 if YoutubeIE.suitable(video_url):
710 entries.append(self.url_result(video_url, 'Youtube'))
713 # here's a fun little line of code for you:
714 video_id = os.path.splitext(video_id)[0]
719 'uploader': video_uploader,
720 'title': video_title,
723 if len(entries) == 1:
726 for num, e in enumerate(entries, start=1):
727 e['title'] = '%s (%d)' % (e['title'], num)